12-gene prognostic signature for breast cancer survival
View Patent ↗A biomarker panel of 12 genes based on expression levels. Methods for calculating prognostic scores and patient ranking based on their score and divided into two equal sized cohorts. Kaplan-Meier analysis and a log-rank test were used to determine differences in survival.
1. A method for calculating a cancer patient's prognostic score comprising steps of: (a) measuring the gene expression level of the EPS15, MELK, NUF2, RNASEH2A, S100P, THYN1, TIMM17A, TSC1, USP47, ZBTB16, PLPP1, and PLEKHH2 genes in a patient's breast tumor tissue with a biomarker panel comprising the EPS15, MELK, NUF2, RNASEH2A, S100P, THYN1, TIMM17A, TSC1, USP47, ZBTB16, PLPP1, and PLEKHH2 genes; (b) calculating the prognostic score using the formula
∑
i
=
1
12
(
gene
i
β
)
×
(
gene
i
expression
level
)
wherein the beta values for each of the 12 genes is obtained from Cox Hazard Regression model; (c) assigning the patient to an appropriate prognostic group based on the calculated prognostic score, whereby a high score above a threshold number indicates the patient has poor prognosis for breast cancer and a low score below the threshold number indicates the patient has good prognosis for relapse-free survival from breast cancer.